This article identifies five Atlantic features in Asian varieties of Creole Portuguese. These reflect common substrate influence from Niger-Congo languages converging with Portuguese forms. This study supports the hypothesis of Dalgado (1917: 41) that the frequent contact between speakers of the different varieties of Asian Creole Portuguese had led to a partial reciprocal diffusion of these creoles’ features, setting this scenario within the larger context of the hypothesis of Clements (2000) that there existed both a general pidgin (spoken in Africa and Asia) and distinctive regional pidgins more influenced by local substrate languages. After examining the possibility that the Malayo-Portuguese feature of marking distributive plurality through noun reduplication may have spread to Indo-Portuguese, this study claims that at least five features of Portuguese-based creoles in Asia had their origin in Africa: (1) the form vai ‘to go’; (2) completive kaba; (3) the coordinating conjunction ku; (4) the preposition na; (5) the negator nunca. This leads to the conclusion that what the Portuguese brought with them to Asia in the 16th century was a general Portuguese pidgin that had been developing in Africa during the second half of the 15th century. This pidgin must have been far more variable and much less developed (i.e. less influenced by substrate languages) than the modern creoles-in all probability a pre-pidgin foreigner talk continuum.